Quality Assurance 5
case it can be observed it is indicative for addition of syrup that was produced
Several indirect methods exist that make use of the presence of certain
enzymes. The idea behind that is to detect enzymes that are used for
production of syrup e.g. from starch, and do not occur naturally in honey.
The concentrations are very low and not easy to analyze directly. The indirect
approach involves the addition of a substrate to honey that is metabolized
only by a certain enzyme that does not occur naturally in honey.
After a short incubation time substrate and metabolite are quantified. If the
metabolite was detected, it proves the presence of the enzyme which is not
supposed to be found in honey.
In the past few years, technology made big leaps forward. It is now
possible to use so-called non-targeted fingerprinting methods, e.g.
NMR-profiling (Nuclear Magnetic Resonance Spectroscopy) or high resolution
mass spectrometry (HRMS), usually coupled to liquid chromatography
(LC-HRMS). These methods often yield huge amounts of data. A single
sample run using an LC-HRMS-system can easily produce half a gigabyte of
data or even more. A single analysis of a honey sample contains thousands
of peaks of different molecules and their fragments. It is obvious that such
amounts of data cannot be evaluated by hand. Instead, increasing computational
power is needed that enables application of statistical techniques,
such as Principal Component Analysis (PCA) or linear discriminant analysis
(LDA) and others to find the needles in the haystack that can be used to
prove or disprove the authenticity of a honey sample.
Syrups that are used for adulteration are usually very well cleaned up
from substances that would make a syrup addition to honey traceable. However,
the high sensitivity of LC-HRMS enables us to detect syrup markers even
in highly cleaned up syrups and thus to prove the presence of syrup in honey.
One might ask if LC-HRMS is really necessary as NMR-profiling is already
a very powerful screening tool that delivers a very comprehensive fingerprint
of a honey sample. It is indeed a legitimate question.
Typically, mostly 400 MHz NMR are used for analysis of honeys. The
huge advantage of NMR lies in the degree to which a method can be
standardized. For example, users of Honey-Profiling™ (an NMR analysis
tool for honey utilizing the worldwide largest honey database (contains
more than 18,000 honeys) offered by Bruker GmbH, Rheinstetten, Germany)
are all using the same sample preparation, same resonance frequency,
same pulse programs and other instrument settings and even the NMR
itself is built by the same manufacturer. Furthermore, NMR has an excellent
reproducibility compared to other methods. That makes NMR a technology
suited for building large databases that can be used independently by other
labs, as long as the same protocol is used.
However, NMR is focusing on the major constituents of honey in the
percent range (e.g. sugars) down to concentrations in the mg/kg range (e.g.
HMF). Especially the sugars have very strong signals in a broad range of the
NMR profile due to their high concentration in honey. Minor compounds that
have signals at the same position are thus completely covered by the strong
sugar signals and are not available for evaluation.
LC-HRMS is about 1000-fold more sensitive than NMR. However, it is not
the method of choice for sugars or other honey constituents in high concentrations.
Instead, LC-HRMS is used particularly to search for (unknown) minor
compounds in the μg/kg range. In particular, we are searching for markers
specific for syrups that do not occur in authentic honey. Thus, NMR and
LC-HRMS are complementary to each other and make a perfect combination.
For LC-HRMS non-targeted as well as targeted analyses are combined.
Non-targeted approaches usually require a database of a sufficient number
of samples to be able to provide statistically sound results.
In order to build up a database, a fixed protocol for sample preparation and
analysis is necessary to ensure that data acquisition was done using the same
Table 1: Overview of samples and analytical results from
conditions. An example is NMR Honey Profiling™. The same sample protocol
is always applied and the same type of NMR is used and also data processing
is standardized. This made it possible to provide a database which can be
used by all NMR Honey Profiling™ users. However, the situation for LC-HRMS
is different. First of all, there are many different manufacturers which use different
technologies. Secondly, neither sample preparation or chromatographic
column and gradients nor mass spectrometer settings are standardized. Every
lab enjoys a maximum degree of freedom for method development.
Thus the obvious question is, if results of an LC-HRMS method developed
in one lab will lead to the same conclusion as results of an LC-HRMS method
developed in another lab.
The two laboratories FoodQS in Langenzenn, Germany, and Quality Services
International (QSI) together with Tentamus Center for Food Fraud (TCF²),
Bremen, Germany, got to the bottom of this question by conducting interlaboratory